In 1971, McCabe et al. 1 conducted a blind family study of a group consisting of 28 good-prognosis and 25 poor-prognosis schizophrenics. The results of that study showed that the families of poor-prognosis schizophrenics contained significantly more schizophrenia, neurosis, and overall illness, but significantly less affective disorder, than the families of good-prognosis schizophrenics. On the basis of their study, the data seemed to support the idea of at least two illnesses in schizophrenia: namely, that which could be called “good prognosis,” which tends toward affective disorder, and that which could be called “poor prognosis,” which tends toward process schizophrenia. Subsequently, Fowler et al. 2 reported a similar blind family study of a group consisting of 35 additional poor-prognosis schizophrenics and the 25 poor-prognosis schizophrenics from the McCabe study, 1 for a total of 60. Probands and their families were classified and then compared, according to paranoid (32) and nonparanoid (28) subgroups, to determine if there are still further distinct illnesses in schizophrenia on the basis of familial data. Family differences between the two groups were limited to a greater risk for the combination of schizophrenia, undiagnosed psychosis, and paranoid personality in the nonparanoid probands' families. This paper presents the results from the further analysis of data gathered for the 60 probands and their families. In addition to the overall frequency of schizophrenia in this group, frequency is calculated for the combined variables of sex, relationship to proband, and diagnostic subtype (of both the proband and of the ill family member).